Large-scale computing architectures made up of computing elements include data storage architecture made up of data storage elements. For example, a number of virtual machines running on a computing device may have virtual file systems with logical data paths, or logical storage volumes, that are implemented by a storage subsystem of the computing device. The storage subsystem of the computing device may itself include a pool of logical storage volumes. As such, the data path of each virtual machine may be implemented on one of the logical storage volumes of the pool of logical storage volumes within the storage subsystem.
The storage subsystem of the computing device may itself be implemented using a storage-array or a storage-area network (SAN), a network-attached storage (NAS), or another type of storage subsystem. In the case of a SAN, the SAN may include a number of logical disks having logical unit numbers (LUN's), where each logical disk is implemented by a physical volume like a physical storage device such as a hard disk drive. Each logical storage volume of the pool of logical storage volumes within the storage subsystem of the computing device may be implemented by multiple physical storage devices. The resulting data storage architecture is thus quite complex, with logical data paths mapping to logical storage volumes, which themselves are mapped to physical storage devices of logical disks of a SAN.
As such, visualizing and analyzing such a data storage architecture can be difficult. This means that undesired redundancies within the data storage architecture may escape detection, such that efficient utilization of data storage resources is not realized. For example, a redundant array of independent disks (RAID) scheme may be employed both in software on the storage subsystem of the computing device, as well as in hardware on the SAN. Such redundant utilization of RAID, however, means that the actual physical storage devices of the SAN are not efficiently employed in an optimal manner, and such inefficient utilization of the SRN's physical storage devices may not be able to be detected by administrators of this data storage architecture.